Understanding Drawdown: What That Percentage Actually Measures
Most traders read drawdown as "how much I lost." That is close enough to be useful and wrong enough to cause bad decisions. The number is a ratio, and ratios depend entirely on what sits underneath them.
This page explains exactly what the drawdown figure on a Freya bot or a marketplace listing is measuring, why the same trading outcome can display two different percentages, and what the metric hides in a strategy that never takes a loss on purpose.
Key Takeaways
- Drawdown measures the drop from your equity's highest point, where equity means allocated capital plus profit or loss, not profit alone.
- Because capital is in the denominator, the same dollar loss shows a smaller percentage on a larger allocation. Two bots with identical trades can display different drawdowns.
- Drawdown is measured from the peak, so a bot has to make money before it can show a meaningful drawdown. A brand new bot showing 0% has not proven anything.
- A DCA grid without a stop loss tends to show low drawdown for a long time, because an open losing position is not a realised loss. The risk is deferred, not absent.
- Read drawdown next to the allocation it was measured against. On its own the percentage is not comparable between bots.
What the number is
Drawdown answers one question: from its best moment so far, how far down has this bot been?
Two definitions matter, and the second is the one people miss.
Equity is your allocated capital plus the bot's profit or loss at that moment. If you allocated 1,000 USDT and the bot is up 50, equity is 1,050. If it is down 80, equity is 920.
Peak equity is the highest that figure has ever reached during the period being measured. It only moves up, never down.
Drawdown at any moment is the gap between the two, as a share of the peak:
drawdown % = (peak equity - current equity) / peak equity × 100
The figure shown as "max drawdown" is simply the largest value that expression reached over the whole period.
A worked example, using an allocation of 1,000 USDT:
| Moment | Profit | Equity | Peak | Drawdown |
|---|---|---|---|---|
| Start | 0 | 1,000 | 1,000 | 0% |
| Good run | +200 | 1,200 | 1,200 | 0% |
| Pullback | +50 | 1,050 | 1,200 | 12.5% |
| Recovery | +260 | 1,260 | 1,260 | 0% |
Notice the pullback row. The bot is still up 50 USDT overall, and it is showing a 12.5% drawdown. That is not a contradiction. Drawdown is not asking whether you are in profit, it is asking how much you gave back from the best point.
Why the same loss shows different percentages
This is the part that makes drawdown non-comparable between bots, and it follows directly from capital being in the denominator.
Take one bot that goes 200 USDT underwater at its worst point:
| Allocated capital | Peak equity | Worst equity | Drawdown shown |
|---|---|---|---|
| 500 USDT | 500 | 300 | 40% |
| 1,000 USDT | 1,000 | 800 | 20% |
| 5,000 USDT | 5,000 | 4,800 | 4% |
Same trades. Same 200 USDT. Three very different headline numbers.
So when you compare two listings and one shows 8% while the other shows 25%, you have learned nothing yet. The second bot may be trading identically on a smaller allocation. Always read drawdown together with the capital it was measured against.
There is an edge case worth knowing about. If no capital figure is available, the calculation falls back to measuring against profit alone. When that happens the percentage can look enormous even for a small loss, because the denominator is tiny. A drawdown figure that looks implausibly large usually means the capital context is missing, not that the bot collapsed.
Why a new bot shows 0%
Drawdown is measured from a peak, and at the start the peak is your starting capital. A bot that has never been in profit and is currently down has a drawdown, but a bot that has barely traded has almost no history to have fallen from.
That leads to a trap when comparing strategies: a low drawdown on a short track record is not evidence of safety. It usually means the bot has not lived through a bad stretch yet. A drawdown figure gets more meaningful the more market conditions it has survived, which is why the period matters as much as the number.
What a stopless DCA grid hides
Now the part that matters most, and the reason this metric can flatter a strategy.
A DCA grid buys more as price falls. Done without a stop loss, the strategy has no mechanism that turns a bad position into a realised loss. It simply holds and adds. The position stays open, waiting for a bounce that usually comes.
Here is the honest version of what that means for the number:
- Unrealised losses do count in drawdown, because equity includes the current value of what is open. A deeply underwater grid will show it.
- But the strategy rarely closes at the bottom, so the recovery erases the dip from the equity curve, and the shape you are left with is a series of small green closes.
- Win rate stays high because closed trades are almost all winners. The losers are still open.
None of this is a flaw in the metric. It is a flaw in reading one number in isolation. The risk in a stopless grid is not that it loses often. It is that it is deferring an outcome, and the deferral has a limit: the capital available for further safety orders.
The practical question is not "what is this bot's drawdown?" It is: if price keeps going against this position, at what point does the strategy run out of room? That is a question about your ladder depth and remaining capital, and drawdown does not answer it.
Two things worth checking on any DCA bot, which the percentage will not tell you:
- How far the ladder reaches. If your safety orders cover a 15% move and the market moves 30%, the strategy has nothing left to do but hold.
- What share of your capital one bot can consume. A grid at full extension is not diversified, it is concentrated at the worst possible time.
How to read it in practice
- Compare drawdown only between bots with similar allocations, or convert back to a currency amount before comparing.
- Read the period alongside the number. 5% over three weeks and 5% over eight months are not the same claim.
- On DCA strategies, look at ladder depth, not just drawdown. The metric describes what happened; ladder depth describes what can still happen.
- Treat a 0% drawdown as missing information, not as a green light.
Drawdown is a good metric that is easy to misread. It tells you honestly what a bot has been through. It says nothing about what it has not been through yet, and with strategies designed to avoid realising losses, that gap is exactly where the risk lives.